Destructive Attribute MSA for Simple Water Leak Check in Trailers

[Swoop4321]

Hi everyone! I'm new here, I've been searching high and low for an answer to my MSA problem and haven't been able to find one, but you all seem very knowledgeable so I thought I'd sign up and see if one of you can help.

Like the subject says, I'm looking for a destructive ATRIBUTE gage study method. I found several threads here with a search, but they all ended without being answered.

My study is for a simple water leak check in trailers where it either gets water inside (fails) or doesn't (passes). However, due to the nature of the process, each trailer absolutely cannot be checked twice. The associate checks it, and if it leaked, they repair it or offline it for repairs. Luckily, every trailer is exactly the same so batches can be made where they are assumed to be the same part (nested R&R) but I just haven't been able to find how to do the study with only pass/fail data.

This has been driving me nuts for weeks and so I would GREATLY appreciate any help I can get!

 

[harry]

Welcome to the Cove

Check Miner's Blog

 

[Swoop4321]

Thanks for the welcome Harry!

I have checked his blog and while it describes standard attribute agreement analysis as well as destructive R&R on variable data, it doesn't speak of any method for destructive R&R on attribute data.

 

[harry]

...................... I have checked his blog and while it describes standard attribute agreement analysis as well as destructive R&R on variable data, it doesn't speak of any method for destructive R&R on attribute data.

I am sorry. Pending further response from Miner or others, you may want to check some of our old threads - destructive R&R on attribute data.

 

[p_tww]

maybe help, if you could prepare batched sample you treated same, Try hypothesis test , but are you sure the sample could be treated as same?

 

[Jim Wynne]

Thanks for the welcome Harry!

I have checked his blog and while it describes standard attribute agreement analysis as well as destructive R&R on variable data, it doesn't speak of any method for destructive R&R on attribute data.

Before you can do an attributes test, you have to have a suitable number of parts, and some of them have to be known leakers. If you're trying to verify that the system will consistently differentiate between good and bad, there's no other alternative.

Why do you feel that MSA would be beneficial?